University of Groningen
Klotho in vascular biology
Mencke, Rik
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Mencke, R. (2018). Klotho in vascular biology. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
401
Chapter 13
Clinical aspects of Klotho in vascular disease
R. Mencke J.L. Hillebrands
402
Abstract
Klotho is a circulating kidney-derived anti-ageing protein with profound effects on arterial health. Klotho deficiency in mice induces a host of vascular pathologies, including vascular calcification, endothelial dysfunction, intima hyperplasia, arterial stiffening, arteriolar hyalinosis, and hypertension. Aside from potential therapeutic uses, it has been investigated whether Klotho would be suitable as a biomarker for vascular disease. Therefore, we discuss studies that have studied Klotho gene variants, Klotho promoter epigenetic modifications, Klotho mRNA expression, Klotho protein expression, and serum Klotho levels, with regard to vascular disease. The functional gene variant KL-VS may be a suitable risk factor for cardiovascular disease, as KL-VS homozygosity is associated with a higher prevalence of CAD and heterozygosity may be associated with favourable outcomes. More studies are to be performed, because in some populations a different effect is observed. Similarly, the A allele of the G-395A polymorphism may be a risk factor for cardiovascular disease. Klotho methylation has been little studied and vascular Klotho mRNA levels are likely too low at baseline to reliably detect a decrease. Vascular Klotho protein expression is a much disputed topic and the likely absence of a well-characterized Klotho protein in arteries would currently preclude a role as a biomarker. It would be interesting to address whether PBMC Klotho methylation or renal methylation, or gene/protein expression levels predict vascular disease. Finally, while many studies have tried to link soluble Klotho levels to vascular disease, the current commercially available ELISAs are not reliable enough to draw meaningful conclusions from these assorted studies. Further studies on Klotho gene variants, the development of a reliable ELISA and the replication of clinical studies on Klotho levels will likely prove to be worthwhile in defining a role for Klotho as a biomarker in vascular disease.
403
Introduction
Klotho is considered an anti-ageing protein, owing to its effects on lifespan and ageing-related diseases. In mice, deficiency of Klotho results in a phenotype reminiscent of human ageing (1) and overexpression of Klotho leads to extension of lifespan by 20-30% and confers protection against the development of many ageing-related diseases (2). Klotho is predominantly expressed in the renal distal convoluted tubule as a membrane-bound protein, which forms a ternary complex with FGFR1c and FGF23, inducing phosphaturia (3). Soluble Klotho can be found in blood and urine after shedding of the membrane-bound protein (4). Many of the systemic effects of Klotho can be attributed to soluble Klotho in the blood (5, 6).
Chronic kidney disease (CKD) is a state of Klotho deficiency (7). Similar to Klotho-deficient mice, CKD patients develop marked vascular calcification. Cardiovascular disease (CVD) is the most common cause of death for CKD patients (8), suggesting that the role of Klotho in the development of vascular calcification may be of particular importance. Klotho may also have broader effects on the vasculature, since various degrees of Klotho deficiency are accompanied by endothelial dysfunction, intima hyperplasia, arterial stiffening, hypertension, impaired angiogenesis, and arteriolar hyalinosis, all of which are ageing-related vascular pathologies (1, 9-14).
In light of these recent developments, it has also been investigated whether Klotho may serve as a biomarker for vascular disease. In this review, we will discuss the studies that have linked the genetics, epigenetics, mRNA expression, tissue expression, and serum levels of Klotho to vascular disease and any technical impediments that exist, to assess whether a role as a biomarker may be envisioned for Klotho.
Klotho gene variants and vascular disease
KL-VS
A number of variants of the human Klotho gene has been identified, the most interesting of which is the KL-VS variant, in which six SNPs are in complete linkage disequilibrium, of which two SNPs (rs9536314 and rs9527025) result in amino acid substitutions (F352V and C370S, respectively) and likely decreased Klotho dimerization, and increased FGF23/FGFR1c binding and signaling (15). In Caucasian and African-American populations, this Klotho variant is prevalent at an allele frequency of around 0.16 (16-19) and it has not or rarely been detected in Iranian and Korean populations (20-23). It has been found that heterozygosity for KL-VS leads to a higher life expectancy (16, 24, 25) and better cognitive function (26, 27), whereas
404
KL-VS homozygotes have a shorter lifespan (16, 24) and worse cognitive function (19, 27). These effects, however, are not found in every study (28, 29).
The first study on the KL-VS variant and vascular disease was performed by Arking et al., and identified that KL-VS prevalence appeared to have an additive effect on the presence of occult coronary artery disease (CAD) in predominantly white and apparently healthy siblings of CAD patients < 60 years of age (OR 1.90 [1.21-2.98]) (17). The effect was mitigated by concomitant hypertension and high HDL cholesterol levels, while smoking unmasked the detrimental effects. This modification by other factors may explain why in a second cohort, which was exclusively African-American and in which patients more often had hypertension and higher HDL levels, no independent effect of KL-VS was found. In this cohort, however, normotensive KL-VS carriers with low HDL levels (<40 mg/dL), also had a higher risk of occult CAD (OR 3.8 [1.29/10.9]). While this study did find the negative effects of KL-VS homozygosity on CAD, the reported life expectancy benefit of KL-VS heterozygosity was not reflected in the CAD risk. A follow-up study in Ashkenazi Jews, however, did find this differential effect of KL-VS status, with significantly fewer cerebrovascular events being reported for KL-VS heterozygotes, while KL-VS homozygotes had experienced more events (24). There was, however, no difference with regard to the incidence of myocardial infarction. Assessing their cardiovascular risk profile, KL-VS heterozygotes had a 6.5 mmHg lower systolic blood pressure than WT allele carriers, while KL-VS homozygotes had a 19 mmHg higher systolic blood pressure than heterozygotes. LDL and HDL cholesterol showed similar patterns, with a favourable profile in heterozygotes and an unfavourable profile in homozygotes. Not every study, however, identifies a similar effect of KL-VS genotype on the presence of coronary artery disease. For instance, although not their primary interest, Deary et al. do not detect a difference in the incidence of cardiovascular disease according to KL-VS allele status in a general population cohort (19). A study by Allon et al. also does not identify an association between KL-VS (heterozygotes + homozygotes vs wild-type) and arteriovenous graft failure, which reflects processes such as thrombosis and intima hyperplasia (30). Whereas the studies by Arking et al. assessed whether individuals with a certain genotype (FF, FV, or VV, referring to the F352V amino acid substitution) had a higher risk of cardiovascular disease than those with a different allele status, a different approach is to assess whether a CVD population is associated with a certain genotype more or less than a control population. Using the latter case-control design, Low et al. do not detect a difference in genotype frequency when comparing patients with premature CAD and controls, although KL-VS heterozygotes were less frequent in the CAD group (24.8% vs 30.8%) (18). Depending on the incidence or prevalence of CVD in a population and given the KL-VS allele frequency, either approach can be expected to encounter power problems. For example, Arking et al. investigated siblings of CAD patients under 60, which would reasonably be expected to have an increased risk of having occult CAD themselves, which indeed resulted in a 23% cohort-wide diagnosis of occult CAD. An attempt, however, to assess whether KL-VS homozygotes are more common in CAD patients as compared to controls will require a large cohort, as KL-VS homozygosity can only be found in around 2% of individuals in most populations.
405
The effect on ischemic cerebrovascular events described by Arking et al. was corroborated by a study by Majumdar et al., who addressed this association between KL-VS allele status and ischemic stroke in an Indian population (31). They found that the wild-type FF genotype conferred a risk (OR 1.48 [1.10-2.00]) compared to the FV genotype, which was modified by age, sex, and smoking. No effect of KL-VS homozygosity on stroke was found in the whole sample, but in stroke patients < 40 years of age, the VV genotype did significantly confer a higher risk for ischemic stroke (OR 3.16 [1.06-9.39]). In the control population, KL-VS homozygotes also had a significantly higher systolic blood pressure. This finding was independently confirmed again by the same group in a different cohort, showing that in a selection of 127 metabolic syndrome patients, those with the VV genotype had a higher systolic blood pressure compared to the other genotypes (32). Conversely, in the larger, unselected cohort, KL-VS homozygosity increased the risk of both hypertension (OR 5.21 [1.00-38.43]) and metabolic syndrome (OR 15.88 [2.56-98.70]), both of which are quite relevant to the risk of cardiovascular disease. In a third study, in a healthy South-Indian population, KL-VS homozygosity was again found to increase the risk of a high systolic blood pressure (33). With regard to other cardiovascular risk factors, however, they and others did not find associations between KL-VS genotype and diabetes, parameters of insulin resistance, or dyslipidemia (31-34). Interestingly, Nzietchueng et al. report rather contrary results in a large, French cohort, finding that KL-VS homozygosity was actually associated with a lower systolic blood pressure and pulse pressure, with KL-VS homozygosity interacting with anti-hypertensive treatment, leading to lower SBP and PP (35). It is unclear why these results are so discrepant from other studies in Caucasian populations, but it is possible that the effects of the VV genotype on blood pressure and the development of macrovascular events are partially independent, or that the effect is modified by environmental factors, other genetic factors, and other determinants. Similarly, a large subset of the Framingham Offspring cohort with data from 3528 subjects did not reveal an association between KL-VS genotype and the presence of aortic and valvular calcifications (36). It is interesting to note that the effects reported are in the same direction as in most studies (lower risk of calcification in the FV group (OR 0.93 [0.81-1.08]) and higher risk of calcification in the VV group (OR 1.36 [0.90-2.06]). A study by Donate-Correa et al. in a patient population that underwent coronary artery bypass grafting or valve replacement surgery, similarly found that atherosclerosis but not valvular calcification was present more commonly in VV + FV patients, as compared to FF patients (37). Although they did not assess the VV and FV genotypes separately, they do note that all VV patients had coronary artery disease, as compared to 71% and 62% of FV and FF patients, respectively, indicating that the VV genotype were in fact most severely affected. The finding that 12 out of 105 patients from this cardiothoracic surgery population had a VV genotype (with a minor allele frequency of 0.31) also indicates either an increased KL-VS allele frequency in Tenerife or an association with the selected population. With regard to the apparently discrepant findings when
406 Ta bl e 1. Th e K L-VS Klo th o gen e vari an t i n vasc ul ar di sease . Re fer en ce Mi no r alle le freq uen cy N Eth ni city Po pul at ion Pa th ol ogy OR Stu dy co ncl usi on Arki ng et al ., 2003 0.153 520 Predo m in an tly Cau cas ian Ap par en tly h eal th y sib lings o f p at ie nts ho sp ita lize d wi th CA D at <60 y ear s o f a ge Occu lt CA D 1.90 [ 1. 21 -2.98 ] fo r FV+ VV v s FF Mo re CAD i n co m bi ne d FV + VV v s FF grou p Arki ng et al ., 2003 0.164 436 Af rican -Am eri can Ap par en tly h eal th y sib lings o f p at ie nts ho sp ita lize d wi th CA D at <60 y ear s o f a ge Occu lt CA D No t s ho w n, ef fe ct m iti gat ed b y hype rt en sio n a nd HDL. 3.8 [ 1. 29 -10.9] in no rm oten siv e in div id ua ls w ith HDL < 40 m g/d L (FV +VV gen ot yp es) No t m ore CAD i n co m bi ne d FV + VV vs FF grou p wi th m ore h yp erten sion a nd hi gher H DL le ve ls Arki ng et al ., 2005 0.262 525 As hken az i J ew s Ve ry o ld : 216 in di vi du al s o ve r 95 yea rs o f a ge Stro ke, MI Stro ke: 5.88 [ 1. 18 -29.41 ] fo r FF v s FV, 30 .65 [2.55 -36 8.00 ] for VV v s FV. MI: n o d iff er en ce Mo re stro ke i n V V gro up , le ss in FV grou p. MI in ciden ce s a re si m ilar. De ar y et al ., 2004 0.173 451 Scott ish Gen er al p op ul at io n CVD No d iffer en ce No d iffer en ce i n C VD occurr en ce b et w ee n ge no ty pe s Lo w et al ., 2005 0.141 (CAD) 0.182 (con tro l) 294 (CAD) 143 (con tro l) No t d iscl os ed Prem at ure CA D pat ie nts v s co ntro ls Prem at ure CA D No d iffer en ce No t m ore VV or f ew er FV carri er s i n p at ie nts wi th CA D comp ar ed to co ntro ls
407 Tav ak ko ly -Baz za z et al ., 2011 0.000 107 Ira ni an CA D p at ie nts a nd con tro ls Sy m pto m at ic C AD - No K L-VS i n t he Ira ni an po pul at io n Rh ee et al ., 200 6 0.004 274 Korean Pa tient s wi th ch es t pai n CAD - KL -VS i s h ar dl y p re se nt in th e Ko re an p op ul at ion Ma ju m dar et al ., 2010 0.123 (stro ke) 0.152 (con tro l) 460 (stro ke) 574 (con tro l) Sou th -In di an Isch em ic s tro ke pat ie nts v s co ntro ls Isch em ic s tro ke 1.48 [ 1. 10 -2.00 ] fo r FF v s FV 1.70 [ 0. 73 -3.94 ] fo r VV v s FV FV ge no ty pe p ro tect s aga in st isch em ic stro ke. V V is a ss oci ated w ith h igher SB P. Ma ju m dar et al ., 201 1 0.157 428 Sou th -In di an Gen er al p op ul at io n (29 % w ith me ta bo lic sy nd ro me, 4 3% hype rt en sive ) Hy pe rten sion , m et abo lic syndr om e Hy pe rten sion : 5.21 [ 1. 00 -38.43 ] fo r VV v s FF M et ab ol ic sy nd ro m e: 1 5.88 [2.56 -98. 70] fo r VV v s FF VV i s a ss oci at ed wi th hy pe rten sion an d me ta bo lic syndr om e Ma ju m dar et al ., 2011b 0.157 429 Sou th -In di an Gen er al p op ul at io n Hy pe rten sion , dy slip id em ia - VV i s a ss oci at ed wi th h igher SB P, n o d iff er en ce in lip id le ve ls Fre at hy et al ., 2006 0.153 (D M) 0.158 (con tro l) 1793 (DM) 1619 (con tro l) Cau cas ian (U K) Ty pe 2 di ab eti cs , con tro ls Di ab ete s, i ns ul in re sist an ce, dy slip id em ia DM: 0.98 [0.8 4-1.14] fo r FV v s F F 0.86 [ 0. 58 -1.28 ] fo r VV v s FF No d iffer en ce i n D M occurr en ce (o r r el at ed par ame ter s) b et w ee n ge no ty pe s Nzi etch ue ng et al ., 2011 0.162 1145 Fre nch Gen er al p op ul at io n SB P, DBP, P P - VV i s a ss oci at ed wi th lo w er SB P an d P P Tan gri et a l., 2011 0.188 3528 Cau cas ian Gen er al p op ul at ion Aort ic a nd a orti c v al ve cal cif icati on 0.93 [ 0. 81 -1.08 ] fo r FV v s FF 1.36 [ 0. 90 -2.06 ] fo r VV v s FF No ef fe ct of K L-VS o n vasc ul ar/ val vu lar cal cif icati on
408 Don at e-Corr ea e t al ., 2015 0.310 105 No t d iscl os ed CA BG a nd AVR pat ie nts Aort ic s ten os is, ca rd iac val ve ca lci ficati on , CA D, at he ro sc le ro sis - VV i s a ss oci at ed wi th CA D an d a th ero scl ero sis , n ot w ith ca lci ficat ion Al lon et al ., 2012 0.188 350 No t d iscl os ed ESRD p at ie nts AVG fai lu re (th ro m bo sis , ste no sis ) 1.13 [ 0. 88 -1.43 ] fo r FV+ VV v s FF No ef fe ct of KL -VS o n AV G fai lu re CAD , c or on ary ar ter y d isea se; HD L, h ig h-den sit y l ip op rot ei n; M I, myocard ial in farc tion s; C VD , c ard iovascu lar di sease ; SB P, sy st ol ic b loo d p re ss ure ; DM , d ia be tes mel lit us; D BP, di as tol ic b lo od pre ss ure , PP , p ul se p re ssu re; C AB G, cor on ary ar ter y b ypa ss gra fti ng; AV R, ao rt ic val ve re pl ace ment; ESRD , e nd -s tage re nal d isease ; AV G, art er iove no us graf t.
409
assessing associations with KL-VS and coronary artery disease and vascular calcification, it should be noted that vascular calcification, although a very important and independent risk factor for cardiovascular events, reflects a process largely distinct from atherosclerosis. Even though Klotho is considered to be relevant for both and the evidence decidedly supports a role for Klotho in the development of vascular and valvular calcification, Klotho is a pleiotropic protein with highly divergent functions that rely on different molecular mechanisms and on different domains in the protein. It is possible that KL-VS, giving rise to two amino acid substitutions in exon 2, mostly affects the molecular actions that reverberate through certain pathological processes, but not others, or even have a compensatory effect on other processes. In short, there may be a protective effect of heterozygosity for KL-VS on CAD, homozygosity for KL-VS promotes CAD, stroke, and metabolic syndrome, while the reports on KL-VS and blood pressure are conflicting and no effect has been found on vascular / valvular calcification or arteriovenous graft failure (summarized in Table 1). Structure-function analysis of Klotho may eventually explain the discrepancies between certain studies.
G-395A
A second interesting SNP in the Klotho gene is G-395A (rs1207568), which is situated in the promoter region and would therefore be expected to affect transcription and gene expression levels, although the few analyses that have been performed to test this hypothesis have yielded conflicting results. Kawano et al. performed an electrophoretic mobility assay, assessing the binding between G-395A-containing oligonucleotides and nuclear protein extracts from HEK293 cells, and found that the G allele more readily facilitated the formation of a DNA-protein complex than did the A allele, suggesting better binding of transcription factors, although luciferase assay results did not indicate a difference in promoter activity (38). Comparing GG to GA + AA subject groups with regard to vascular Klotho expression levels, Donate-Correa et al. show that the GG genotype is associated with higher Klotho expression levels (37), although it should be kept in mind that arterial Klotho expression levels are extremely low and it would be more reliable to compare renal expression levels. Conversely, Wang et al. also performed luciferase assays and their results indicate that the A allele may be associated with higher promoter activity, although these results were not statistically significant (39). The effect of the G-395A substitution therefore remains unclear.
A number of studies has addressed whether the G-395A SNP results in effects on the development of vascular diseases. In a case-control study, Imamura et al. found that carriers of the G-395A SNP (with genotypes GA or AA) had a higher risk of CAD (OR 1.82 [1.21-2.74]), but not vasospastic angina (OR 1.30 [0.72-2.35]) as compared to individuals with the GG genotype. A study on Korean patients with chest pain (43% CAD) by Rhee et al. reported a minor allele frequency similar to the Japanese CAD patient population, but found no difference in CAD according to G-395A genotype, except for a protective effect of the A allele in patients over 60 (22). SBP, DBP and serum lipid measurements did not reveal differences
410
between GG and GA + AA genotypes. In a second study, however, in healthy subjects, they did find that female Korean A allele carriers had a higher systolic blood pressure than GG homozygotes (40). A third Korean study, an association with CAD was found, with possession of an A allele presenting an increased risk (OR 1.71 [1.07-2.75]) even after correction for traditional cardiovascular risk factors (41). Perhaps similar to the studies on KL-VS and calcification, the A allele of the G-395A was not a risk factor for coronary artery calcification in this study. In a meta-analysis, incorporating the previous three CAD studies and three additional Chinese CAD studies, the overall conclusion was that the A allele indeed conferred a risk of CAD (OR 1.24 [1.06-1.45]), in a total of 1560 CAD patients and 1459 control patients (42). Finally, Donate-Correa et al. did not find a difference between GA or AA carriers and GG carriers, with regard to valvular calcification, hypertension, atherosclerosis, coronary artery disease, dyslipidemia, or left ventricular hypertrophy, while diabetics were more common in the GA + AA group (37). The relatively high minor allele frequency of 0.348 in this predominantly CAD-affected population stands out.
In another Korean cohort, it was examined whether an association exists between ischemic stroke (N=273) or vascular dementia (N=207) and the G-395A polymorphism, which was not the case in neither dominant or recessive models (23). In subgroup analyses, no effects were found either of the A allele on large artery atherosclerosis or small vessel occlusion as cause of ischemic stroke. However, in a subgroup of cardioembolic ischemic stroke (N=20), it was found that the A allele conferred an increased risk, especially in the AA genotype (OR 6.52 [1.68-25.25] for AA vs GA + GG), which was found to be mostly the case in women (OR 9.07 [1.62-50.76]), rather than men. It should be noted that this difference is caused by the presence of two (out of 10) women in the cardioembolism group, as compared to one (out of 10) men, being carriers of the AA genotype. Given that it is striking that the A allele frequency was 0.325 in this CE group of 20 patients, as compared to the overall frequency of 0.183 in the ischemic stroke group of 273 patients, it may indicate an effect of this polymorphism on the pathogenesis of thrombosis or atrial fibrillation, both of which are expected to be affected by Klotho (43, 44). Replication of this study in a larger cohort is warranted.
While KL-VS has so far not been shown to affect arteriovenous graft failure in hemodialysis patients, A allele carriers of the G-395A SNP were found to be overrepresented in the early dysfunction group (45.7% (GA + AA) vs 19.8% (GG)) (45). This can be largely attributed to the contribution of the 3 AA genotype carriers (75%), as the GA genotype contained an equal 19.8% early vascular access failure patients. Nevertheless, within a period of 12 months, A allele carrier AV fistulas did have a significant survival disadvantage.
With regard to cardiovascular risk factors, two studies by Shimoyama et al. have addressed whether the A allele is associated with a high-risk profile. In men, it was shown that A allele carriers have lower HDL cholesterol levels, whereas in women, the A allele was associated with higher fasting glucose levels, but systolic blood pressure was not significantly different (46). In hemodialysis patients LDL cholesterol levels were not different with regard to G-395A
411
genotype and while coronary artery calcification appeared to be lower in female A allele carriers, this difference disappeared after correction for age and duration on hemodialysis (47). Ko et al., in a Korean population, reported that A allele carrier status was associated with a higher prevalence of hypertension in an IgA nephropathy cohort (48) and of diabetes in a hemodialysis patient cohort (49), whereas in neither cohort an association was found with serum lipid levels. Wang et al. actually report a protective effect of the A allele on the development of essential hypertension, even after correction for traditional cardiovascular risk factors (OR 0.59 [0.38-0.94]) (39). A larger Chinese study in 710 test subjects, however, reported similar results (adjusted OR 0.68 [0.49-0.95]), also with a significantly lower SBP in A allele carriers (50). It should be noted, however, that the minor allele frequencies of 0.218 and 0.200, respectively, in the control populations of these studies are higher than described in other studies, albeit in other populations. In the same cohort, it was found that the A allele is associated with a lower prevalence of metabolic syndrome, hypertension, hypertriglyceridemia, and, in men, low HDL cholesterol levels (51). Considering the sample size, however, the effects in these studies are not likely to be due to chance; furthermore, other studies in Chinese populations also report beneficial effects of the A allele with regard to cognitive function (52).
Overall, it appears that the A allele of the G-395A SNP is associated with an increase in prevalence of CAD and perhaps cardioembolic ischemic stroke, while it may be associated with a decrease in prevalence of hypertension and the effects on cardiovascular risk factors are still relatively little-studied. The fact that most studies identify the A allele as a deleterious factors makes it tempting to hypothesize that the inhibitory effects on Klotho expression as observed by Kawano et al. and Donate-Correa et al. may be responsible. Given that this SNP is located in the promoter region, differential structure-function relationships between different domains of the Klotho protein and the molecules with which it interacts are unlikely to explain the discrepant effects of the A allele on CAD and hypertension. It is possible that the different environments and genetic make-ups of these populations are responsible. In general, as the relationship between G-395A and vascular disease had only been investigated in Asian populations, it is important to assess whether associations exist in other populations as well, as to ascertain whether an independent effect can be ascribed to this SNP.
C1818T
The third most investigated Klotho polymorphism is C1818T (rs211239), which is a synonymous SNP in exon 4 and has not been found to be in linkage disequilibrium with any other Klotho SNP. It is unknown whether this SNP affects gene transcription, splicing, or is in
412 Ta bl e 2. G -39 5A in th e Klo th o gen e in vasc ul ar d isea se. Re fer en ce Mi no r alle le freq uen cy N Eth ni city Po pu lat ion Pa th ol ogy OR Stu dy co ncl us ion Imam ura et al ., 2006 0.150 (CAD) 0.117 (VSA) 0.097 (con tro l) 197 (C AD) 77 (VSA ) 331 (c on tro l) Jap an es e CA D, VSA, con tro l CA D, VSA CA D: 1.8 2 [ 1.2 1-2.74 ], f or G A + AA vs G G VSA: 1.30 [ 0.7 2-2. 35] , for GA + AA vs G G Th e A ge no ty pe is a ris k fact or fo r CAD , n ot fo r VSA Rh ee et al ., 2006 0.148 274 Korean Pa tient s w ith ch es t p ai n CAD 0.91 [ 0.52 -1.5 7] fo r G A + AA vs G G No effec t o f G -3 95A p ol ym or ph ism on CAD Rh ee et al ., 2006b 0.171 243 Korean He al th y wo m en Hy pe rten sion , dy slip id em ia , in su lin re sis ta nce - SB P w as low er i n G G as comp are d to GA + AA. Jo et al ., 2008 0.146 434 Korean Pa tient s w ith ch es t p ai n CA D, CA C CA D: 1. 71 [ 1.07 -2.75 ] f or GA + AA vs G G CA C: n .s. A al le le is a ri sk f act or fo r CA D, n ot CAC Don at e-Correa et al., 2015 0.348 105 No t di scl os ed CAB G an d AVR pa tie nt s Aort ic s ten os is, card iac v al ve cal cif icati on , CA D, at he ro scl ero sis - A al le le i s n ot ass oci ated w ith vasc ul ar p at ho logi es Ki m et al ., 2006 0.183 (stro ke) 0.176 (V D) 0.156 (con tro l) 273 (s tro ke ) 207 (VD ) 455 (c on tro l) Korean Stro ke, VD , con tro l Stro ke (d ue to LAA, SVO, or CE ), VD Stro ke: 1.4 0 [ 0. 60 -3.29 ], f or AA v s GA + GG 1.23 [ 0.89 -1.7 0] fo r AA + G A vs G G LAA: 1.1 9 [0.33 -4.30 ] f or AA vs GA + GG 0.93 [ 0.57 -1.52 ] f pr AA + G A vs G G No e ffe ct of A al le le o n VD or isch em ic stro ke in ge ne ra l, i ncl ud in g LAA an d SVO. Ad ve rs e ef fe ct on CE .
413 SVO: 0.6 7 [ 0.1 5-3. 04 ] f or AA vs GA + GG 1.39 [ 0.90 -2.1 5] fo r A A + GA vs G G CE : 6.52 [1. 68 -25.25 ] f or AA vs GA + GG 2.50 [ 1/02 -6.15 ] f or AA + G A vs G G VD: 1. 10 [1.4 1-2. 98] f or AA vs GA + GG 1.20 [ 0.84 -1.7 1] fo r AA + G A vs G G Ki m et al ., 2008 0.155 126 Korean Pa tient s w ith an AVG AVG fai lu re - A al le le i s a ris k fact or fo r AV G dys func tio n Sh im oy ama et al ., 2009 0.169 473 Jap an es e Gen er al po pul at io n Dy slip id em ia - Lo w er HDL in A al le le ca rri ers Sh im oy ama et a l., 2009 b 0.153 219 Jap an es e He m od ya lisi s pa tie nt s Dy slip id em ia - No d iffer en ce i n L DL Ko e t al ., 2013 0.203 478 Korean He m od ia ly sis pa tie nt s Dy slip id em ia , di ab etes - No d iff er en ce in lip id le ve ls, h ig he r pre val enc e o f d iab ete s i n A ca rri ers Ko e t al ., 2012 0.170 973 Korean IgA nephr opa thy Hy pe rten sion , dy slip id em ia , di ab etes - Hi gher pre val en ce o f h yp erten sion in A carri ers , n o d iff ere nc es i n pre val enc e o f d iab ete s o r l ipi d le ve ls Wa ng et al ., 2010 0.158 (H T) 0.218 (con tro l) 215 (HT ) 220 (c on tro l) Chi ne se HT pa tie nt s, con tro ls Hy pe rten sion Un ad ju sted : 0.5 9 [0.40 -0.89 ] fo r GA +A A vs GG Ad ju sted 0.59 [0.38 -0.9 4] Pro tecti ve e ffe ct of th e A al le le in hy pe rten sion Gao et al ., 2015 0.137 (H T) 0.200 (con tro l) 432 (HT ) 278 (c on tro l) Chi ne se Gen er al po pul at io n >9 0 y ear s ol d, pre do m in an tly hype rt en sive Hy pe rten sion Un ad ju sted : 0.6 7 [0.48 -0.93 ] fo r G A + A A v s G G Ad ju sted 0.68 [0.49 -0.9 5] Pro tecti ve e ffe ct of th e A al le le in hy pe rten sion
414 Lu o et al ., 2016 0.092 (Me tS) 0.163 (con tro l) 65 (Me tS) 630 (c on tro l) Chi ne se Gen er al po pul at io n >90 y ear s o ld Me ta bo lic syndr om e, dy slip id em ia , hy pe rten sion Me tS: 0.50 [ 0.2 5-0.98 ], f or GA + AA vs GG HT : 0.48 [0. 34 -0.67 ] f or G A + AA v s G G Hy pe rtr igl yce rid em ia: 0 .66 [0.39 -0.9 5] fo r G A + AA vs G G A al le le p ro tect s aga in st Me tS, hype rt en sion , an d hy pe rtr igl yce rid em ia CAD , c or on ary ar ter y di sea se; VSA, vasosp ast ic a ngi na; SB P, sy st ol ic b loo d p re ssu re ; C AC, cor on ary art er y cal cif icat io n; CAB G , c or on ary art er y bypa ss gr af tin g; AV R, a or tic val ve re pl acem en t; V D, vasc ul ar d em en tia; LAA, lar ge ar ter y at her oscl ero sis; S VO, sm al l v essel occ lu sion ; C E, card ioe m bol ism ; A VG, art er iove no us graf t; H DL, h igh d en sit y lip op ro tei n; LD L, low -d en sit y lip op ro tei n; HT, h ype rt en sion ; M et S, met ab ol ic sy nd ro me.
415
likage disequilibrium in a haplotype block with a functional SNP. Nevertheless, if C1818T has consequences for disease, it is useful to discuss the studies that have been performed so far. Rhee et al. found that carriers of the T allele had a significantly lower prevalence of CAD in a cohort of patients with chest pain, which was the case in particular for patients under 60 years of age (22). In haplotype analysis, it was also found that patients with both the G-395A GG genotype and the C1818T CC genotype had the highest risk (OR 2.59 [1.39-4.86]) as compared to GG and T carrier status, while other combinations did not reveal a significant difference. Blood pressure and serum lipid analyses did not reveal differences between CC and CT + TT genotypes. In a second study on cardiovascular risk factors in healthy Korean women, T allele carriers had a higher SBP and a higher fasting plasma glucose, although only the glucose association persisted after correcting for age and BMI and no differences were observed with regard to the lipid spectrum or other parameters of insulin resistance (40). Haplotype analysis in this study, however, found an effect not of the G-395A GG genotype, but of concurrent (G-395A) AA and (C1818T) TT genotypes (subjects having both minor alleles, N=40), exhibiting higher SBP, DBP and fasting plasma glucose levels than their concurrent GG and CC counterparts (N=124). The finding of higher fasting glucose levels (and similar HDL levels) in T allele carriers is corroborated by Shimoyama et al., a difference most pronounced in males under 60 years old (46). A follow-up study also includes an indication that LDL cholesterol may be lower in T allele carriers (while the triglyceride level may be higher, however, not after correction for age and duration on hemodialysis), which stands in contrast with previous studies (47). In Korean IgA nephropathy patients, there was no difference between genotypes with regard to serum lipid levels, while in hemodialysis patients, T allele carriers displayed only lower total cholesterol levels (48, 49).
The study by Kim et al., on ischemic stroke and vascular dementia, did not reveal any association with the different C1818T genotypes in either a dominant or recessive model (23). The same holds true for the study on ischemic stroke and C1818T by Majumdar et al. (31), who also in three studies report there to be no differences in fasting glucose or in blood pressure or lipid parameters (31-33).
Overall, it appears that little evidence exists to support an association between the C1818T polymorphism and outcomes or determinants of vascular disease (summarized in Table 3), although the protective effect of the T allele as reported by Rhee et al. may merit replication in a larger cohort (22).
Other SNPs
The H193R mutation was first described by Ichikawa et al., who reported on a 13-year-old girl who was homozygous for this mutation and presented with severe ectopic calcifications, including in the carotid arteries (53). This mutation in exon 1 was found to destabilize a
416 Ta bl e 3. C 18 18 T i n th e Klo th o gen e in vasc ul ar d isea se. Re fer en ce M in or a lle le freq uen cy N Eth ni city Po pu lat ion Pa th ol ogy OR Stu dy co ncl usi on Rh ee et al ., 2006 0.170 273 Korean Pa tient s w ith ch es t pai n CAD 1.73 [ 1.03 -2.9 1] fo r C C vs CT + T T Pro tecti ve e ffe ct of th e T a lle le Rh ee et al ., 2006b 0.196 237 Korean He al th y wome n Hy pe rten sion , dy slip id em ia , in su lin re sis ta nce - SB P w as h igher in T T as comp ar ed to C T + T T, b ut no t after corre cti on fo r age an d BMI. Hi gher f asti ng p las m a gl uco se in T a lle le ca rri er s. Sh im oy ama et al ., 2009 0.177 473 Jap an es e Gen er al p op ul at io n Dy slip id em ia , in su lin re sis ta nce - No d iff er en ce i n H DL, h igh er fasti ng pl as m a gl uco se le ve ls i n T a lle le ca rri er s Sh im oy ama et al ., 2009b 0.171 219 Jap an es e He m od ya lisi s pa tie nt s Dy slip id em ia - Lo w er LDL in T a llel e c ar rier s Ko e t al ., 2013 0.182 478 Korean He m od ia ly sis pa tie nt s Dy slip id em ia , di ab etes - No d iffer en ce in d ia bet es , l ow er to ta l ch ol es terol le ve ls i n T carri er s Ko e t al ., 2012 0.184 973 Korean IgA n ep hro pat hy Hy pe rten sion , dy slip id em ia , di ab et es - No d iffer en ces Ki m et al ., 2006 0.178 ( stro ke) 0.181 (V D) 0.181 (c on tro l) 273 (stro ke) 207 (VD ) 455 (con tro l) Korean Stro ke, VD, con tro l Stro ke (d ue to LAA, SVO, or CE ), VD Stro ke: 0.78 [ 0.31 -1.92 ], f or T T vs CT + CC 1.00 [ 0.73 -1.3 8] fo r T T + C T v s C C LAA: 0.62 [0. 14 -2.78 ] f or T T v s CT + CC 1.07 [ 0.67 -1.6 9] fp r T T + C T v s CC SVO: 0.5 3 [ 0.1 2-2. 37] f or T T v s CT + CC 0.78 [ 0.49 -1.2 3] fo r T T + C T v s CC No e ffe ct of T al le le o n VD or isch em ic stro ke i n ge ne ra l, in clud in g L AA, SVO, an d CE .
417 CE : 1.54 [0.19 -12. 31 ] f or T T v s CT + C C 1.36 [ 0.54 -3.3 9] fo r T T + C T v s CC VD: 1. 03 [0.41 -2. 56] for TT v s C T + C C 1.00 [ 0.70 -1.4 1] fo r T T + C T v s CC Ma ju m dar et al ., 2010 0.240 ( stro ke) 0.257 (c on tro l) 446 (stro ke) 571 (con tro l) Sou th -In di an Isch em ic s tro ke pat ie nts v s co ntro ls Isch em ic s tro ke 0.91 [ 0.71 -1.1 7] fo r C T + T T v s CC 0.82 [ 0.48 -1.3 9] fo r T T vs CC + CT No d iffer en ce in o cc ur ren ce of isch em ic s tro ke Ma ju m dar et al ., 2011 0.243 420 Sou th -In di an Ge ne ra l po pul at io n (29 % w ith met ab ol ic syndr om e, 4 3% hype rt en sive ) Hy pe rten sion , met ab ol ic syndr om e Me ta bo lic sy nd ro m e: 1. 06 [0.62 -1.7 9] for CT + T T v s C C 0.28 [ 0.07 -1.0 7] fo r T T vs C T + CC No ass oci ati on b etwe en C1 81 8T ge no ty pe a nd m et abo lic syndr om e. No di ffe re nc es i n bl ood pr es sur e, i ns ul in re sis ta nce , o r l ipi d p ar ame te rs . Ma ju m dar et al ., 2011b 0.243 420 Sou th -In di an Gen er al p op ul at io n Hy pe rten sion , dy slip id em ia - VV i s a ss oci at ed wi th h igher S BP CAD , c or on ary ar ter y di sea se; SB P, sy st ol ic b loo d p re ssu re ; B M I, bod y ma ss in dex; HD L, hi gh -d en sit y lip op ro tei n; LD L, l ow -d en sit y l ip op ro tei n; VD , vasc ul ar d eme nt ia; LAA, large art er y at her osc ler osi s; SV O, smal l vessel occ lu sion ; C E, car di oe m bol ism .
418 Ta bl e 4. Ot her K lot ho SN Ps in vasc ul ar d isease . Re fer en ce SN P Mi no r al le le freq uen cy N Eth ni city Po pu lat ion Pa th ol ogy OR Stu dy con clusi on Ich ika w a et al. , 2007 H193R - 1 No t di scl os ed - Caro tid ar tery cal cif icati on - H193R cau se s e cto pi c cal cif icati on , i ncl ud in g in th e ca ro tid a rteri es Ogu ro e t a l., 2010 A371 90 G - 853 (HT ) 1783 (con tro l) Jap an es e Hy pe rten siv e, con tro l Caro tid i nti ma -m ed ia th ickn es s - No a ss oci ati on Ogu ro e t a l., 2010 C4 441 3T - 853 (HT ) 1783 (con tro l) Jap an es e Hy pe rten siv e, con tro l Caro tid i nti ma -m ed ia th ickn es s - No a ss oci ati on Ogu ro e t a l., 2010 C3 882 3T 0.079 (H T) 0.080 (con tro l) 853 (HT ) 1783 (con tro l) Jap an es e Hy pe rten siv e, con tro l Caro tid i nti ma -m ed ia th ickn es s OR 0.35 [ 0.0 4-2.26 ] fo r T T v s CC + C T 0.33 [ 0.0 4-2. 17 ] f or CC v s T T 0.47 [ 0.0 6-3. 23 ] f or TT v s C T T al le le carri ers w ere signi fican tly m ore comm on in H T p at ie nts w ith at he ro scl ero sis , bu t th is ass oc ia tio n d id no t pe rs ist after correcti on f or tra di tio nal ri sk fact ors . Th er e wa s n o ass oci ati on i n con tro l pa tie nt s. Ogu ro e t a l., 2010 T453 69A 0.356 (H T) 0.351 (con tro l) 853 (HT ) 1783 (con tro l) Jap an es e Hy pe rten siv e, con tro l Caro tid i nti ma -m ed ia th ickn es s OR 1.69 [ 1.1 8-2.44 ] fo r T T + T A v s AA 1.22 [ 0.6 7-2. 18 ] f or TT vs A A 1.83 [ 1.2 5-2. 68 ] f or TA v s A A T al le le carri ers w ere signi fican tly m ore comm on in H T p at ie nts w ith at he ro scl ero sis , al so after corre cti on fo r tra di tio nal ri sk fact ors . Th er e wa s n o
419 HT, h ype rt en sion ; N SAID , n on -st er oi dal a nt i-i nf la mmat or y d ru gs; C OX -2 , c ycl o-oxy gen ase -2; A CS, ac ut e cor on ary sy nd ro me. ass oc iat ion i n con tro l pa tie nt s. St. Ger ma in e et al. , 2010 rs 650 439 0.235 (NSAID) 0.163 (n o NSAID) 113 (NSAID s) 332 (n o N SAID s) Can ad a AC S p at ie nt s ex po sed o r une xp os ed t o NSAIDs Card iov as cu lar ri sk s of N SAID i nh ib ito rs NSAIDs : 1. 76 [1.11 -2.80] for GG + CG v s CC COX -2 i nh ib ito rs : 2.35 [ 1.4 0-3. 93 ] f or GG + CG v s CC Th e G a lle le i s ass oci at ed w ith b oth NSAID -rel at ed a nd COX -2 in hib ito r-re lat ed dev el op men t o f A CS
420
putative enzymatically active site in the KL1 domain, as well as impair Klotho protein expression, binding to FGF23/FGFR1 and enabling of FGF23 signaling. In contrast to KL-VS, a functional gene variant that has been linked to coronary artery disease, but not to vascular calcification, the H193R mutation illustrates that a modification in a different region is likely to affect different functions of the Klotho protein. It is unknown how prevalent this mutation is in the general population and whether heterozygotes also experience an increased risk. Oguro et al. have investigated a different set of Klotho SNPs (A37190G (rs7323281) in intron 1, C44413T (here designated rs5644481) in exon 4 (synonymous), rs3752472 (C38823T) in exon 3 (P514S), and rs650439 (A45369T) in intron 4) in a cohort of hypertensive patients and in a cohort of healthy controls (54). They found that A37190G and C44413T were not associated with carotid atherosclerosis (defined as a meant intima-media thickness ≥ 1.1 mm) in either cohort. T allele carriers for the C38823T SNP were more common in hypertensive patients with carotid atherosclerosis, but not in controls and there was no difference after correction for traditional cardiovascular risk factors. T allele carriers of the A45369T polymorphism, however, showed a similar association with carotid atherosclerosis (again, only in hypertensive patients), which did persist after correction for confounders. Future studies may shed more light on this SNP as a cardiovascular risk factor. Finally, St. Germaine et al. used a pharmacogenetic approach in a cohort of acute coronary syndrome patients who were exposed or not to NSAIDs or selective COX-2 inhibitors in an attempt to identify SNPs that may predispose to the development of cardiovascular events as a side effect of celecoxib. They studied Klotho SNP rs211247 and found that the G allele conferred a risk of both NSAID- and COX-2 inhibitor-related ACS (55).
From the previous discussion, it becomes apparent that KL-VS homozygosity variant is currently the most promising candidate as a potential risk factor, while KL-VS heterozygosity may confer protection from cardiovascular disease, although further studies are required to assess the effects in larger cohorts and in different populations. Additional structure-function studies are required to place these associations in a mechanistic context. For G-395A, it is important to ascertain whether an effect exists on the gene transcription level and whether the A allele is a risk factor in other populations. In general, a few promising associations have been observed, but it is too early for a role for genetic analysis of the Klotho gene in creating a risk profile for cardiovascular disease in patients.
Epigenetic modifications of Klotho in vascular disease
Klotho gene expression is regulated, among other processes, by promoter methylation and histone acetylation. Klotho is known to be down-regulated markedly in chronic kidney disease
421
(CKD), mostly under the influence of uremic toxins, like indoxyl sulfate (56). CKD is considered a state of Klotho deficiency and both CKD patients and Klotho-deficient mice are exceptionally prone to the development of cardiovascular disease, especially vascular calcification (7). From a potentially therapeutic point of view, demethylation of Klotho may turn out to be a promising approach. From a diagnostic point of view, it may be possible to formulate a cardiovascular disease risk based on a profile that includes the methylation status of the Klotho gene. Such a study has not been performed so far, although it has been shown that Klotho methylation status in PBMCs correlates to both renal Klotho methylation status and to the CKD stage (57), the severity of which in turn is related to the cardiovascular disease risk. Vascular Klotho methylation has also been shown to be increased in CKD patients (58), making is a possible marker, but not a practical one. If Klotho promoter methylation is to be explored as a marker for early detection of vascular disease or as a risk factor or predictor for cardiovascular events, it would be interesting for the putative association to be explored between PBMC Klotho hypermethylation and vascular calcification and cardiovascular outcomes in CKD patients and in non-CKD patients.
Vascular Klotho mRNA expression in vascular disease
As extensively discussed elsewhere (59), it appears that Klotho mRNA expression is generally around the detection limit or (q)RT-PCR assays in arteries, with many studies detecting a low gene expression level and others not detecting anything. A few studies have addressed whether Klotho is differentially expressed in arteries in vascular disease – most in the context of chronic kidney disease, which predisposes to the development of vascular calcification. In this regard, Donate-Correa et al. have shown in thoracic aorta tissue (obtained during CABG) and Van Venrooij et al. have shown in coronary arteries (from hearts obtained during heart transplantation) that Klotho mRNA expression may be detected in human arterial tissue (60, 61). Follow-up studies indicate that Klotho expression is lower in patients with coronary artery disease and perhaps in diabetes (37, 62). There was, however, no difference with regard to patients who had or did not have valvular calcification on echocardiography. We also detected very low Klotho mRNA expression in kidney donor renal arteries and in kidney graft recipient iliac arteries, but there was no difference between CKD or healthy arteries (63), which Chen et al. did find in ESRD radial arteries as compared to radial arteries obtained from patients undergoing CABG (58). Overall, it may be that arterial Klotho expression, low as it may be in healthy arteries, is lower in CKD patients, which is in line with systemic hypermethylation of the Klotho promoter found in CKD. Given the low expression levels and the impracticalities associated with obtaining vascular tissue, it would be more interesting to explore whether Klotho expression in renal biopsies correlates with vascular disease.
422
Vascular Klotho protein expression in vascular disease
Although the extremely low Klotho mRNA levels in arterial tissue are not particularly controversial, the question whether this expression level leads to relevant protein expression is much disputed. Generally speaking, authors usually either detect a protein that is around 116 kDa (whereas renal Klotho is 130 kDa), or they do not detect a protein. This controversy is essentially reliant on the specificity of antibodies and the use of controls to determine what is real and specific expression. We and others have shown that membrane-bound Klotho, as expressed in the kidney, is not expressed in arteries, nor is functional as such in terms of facilitating FGF23 signaling (63, 64). Lim et al., on the other hand, show data that Klotho protein is expressed (as a 116 kDa protein) in arteries and that its expression is lower in CKD arteries (65), which is similar to data reported by Chen et al. (58) Data from Van Venrooij et al., however, indicate that Klotho is especially expressed in calcifications in coronary arteries (61).
The current controversy with regard to vascular Klotho expression precludes the vascular Klotho protein as a potential biomarker. With regard to Klotho protein expression, it would be more interesting to assess whether renal Klotho protein expression is associated with vascular health.
Soluble Klotho levels in vascular disease
A very attractive option for the potential use of Klotho as a biomarker is the shedding of membrane-bound Klotho to the blood and to the urine, where it can be detected as a soluble protein. Almost all circulating Klotho is derived from the kidney (5, 6), and soluble Klotho is known to greatly impact arterial health, since kidney-specific deletion of Klotho induces the development of vascular calcification as it does in the full knockout mouse (6), and since overexpression of Klotho (mostly in organs other than the kidney) prevents the development of CKD-induced vascular calcification (7).
A potential impediment to the study of soluble Klotho levels is the fact that a sensitive and specific ELISA is a requirement. Several assays have been developed and very different results have been reported. Yamazaki et al. were the first to establish an ELISA for the detection of serum Klotho, which was found to be at a level of 562 ± 146 pg/mL in healthy adults and at a significantly higher level in healthy children (952 ± 282 pg/mL) (66). This ELISA employs antibodies 67G3 and 91F1 and it was good at recovering recombinant human Klotho from plasma and serum. Using serum from the H193R mutation patient with tumoral calcinosis and
423
extensive ectopic and vascular calcification, the serum level in this patient was found to be 337 pg/mL, which was low, but only 1.5 SDs lower than the adult average. Devaraj et al. evaluated a different ELISA and found that in healthy adults, the mean Klotho serum level was 103 ± 95 ng/mL with a median of 76 ng/mL, which is quite a factor higher than in the previous study (67). Counterintuitively, these authors report higher Klotho levels in CKD. The discrepancies between different assays prompted a number of comparison studies. Heijboer et al. found that the aforementioned two assays correlated with an R of only 0.055, in addition to the persistent concentration differences (68). The IBL assay (with antibodies from Yamazaki et al.) performed best in terms of agreement between plasma and serum samples and low inter-assay and intra-assay variation. Pedersen et al. performed a similar analysis, also finding no correlation between measurements with different assays (R = -0.04) (69). With regard to the IBL assay, it was shown that freezing serum or urine samples has a tremendous negative impact on the signal that is measured after thawing (70, 71), whereas sequential immunoprecipitation and immunoblotting indicates that Klotho proteins are not notably degraded and can still be detected (71). Furthermore, assessing whether urine Klotho levels are reliable, the signal as measured by the IBL assay decreased markedly after simulation of storage of urine in the urinary bladder at 37 °C for three hours, making it difficult to measure urinary Klotho levels if not in fresh catheter urine that is immediately processed (70). Given these technical issues, it is difficult to appraise the value of the assorted studies on Klotho serum levels that so far have yielded wildly varying results. It appears that the Klotho-knockout validated IP-IB assay using antibody Sb106 is currently the best method, but it is not suitable for large-scale studies. In the following discussion of studies that have investigated serum Klotho levels with regard to vascular disease, it is advisable to exercise caution with regard to interpreting results.
Semba et al. divided a cohort of largely elderly Italians in tertiles according to Klotho levels and found that higher Klotho levels were associated with a lower prevalence of cardiovascular disease (72). Kitagawa et al. tried to link Klotho levels to different parameters of vascular disease in CKD: flow-mediated dilation (FMD) for endothelial function, ankle-brachial pulse wave velocity (PWV) for arterial stiffness, the maximal intima-media thickness (IMT) for atherosclerosis, and the aortic calcification index (ACI) for vascular calcification (73). Higher Klotho levels were seen in patients with healthier arteries with regard to all four of these parameters, although lower Klotho levels were associated independently only with a higher PWV. In HIV-infected patients exposed to combined antiretroviral therapy, a similar association between higher plasma Klotho levels and a lower mean IMT was found (74). Abdallah et al. describe similar findings with regard to IMT and also find a lower prevalence of CAD in patients with higher Klotho levels (75). Keles et al. report using the Cusabio ELISA (although the values are in the IBL ELISA range and below the stated 0.156 ng/mL detection limit of the Cusabio ELISA) and their results suggest that higher Klotho levels may be associated with higher FMD and coronary flow reserve (CFR), as well as with a lower IMT (76-78). With regard to vascular calcification, multiple authors describe a negative association with Klotho. Buiten et al. find that patients with low Klotho levels more frequently have CAD,
424
although low Klotho levels were found not to be an independent risk factor in this study (79). Cai et al. do find an independent association with moderate to severe calcification of the abdominal aorta, with OR 0.63 [0.41-0.96] per SD Klotho increase (80). Di Lullo et al. show that Klotho levels are negatively correlated with AVC (81). Finally, two small studies provide indications that Klotho levels may be lower in hypertension (82, 83).
Many authors, however, do not find associations between Klotho and vascular disease. Shibata et al. did not find any association between serum Klotho levels and the presence of CVD or related parameters including blood pressure, in cardiology inpatients (84). Karalliedde et al. treated diabetic patients with valsartan/hydrochlorothiazide or amlodipine and detected an increase in serum Klotho levels in the valsartan/hydrochlorothiazide group, which, however, did not correlate with decreases in brachial pulse pressure or aortic PWV (85). Seiler et al. report that high FGF23 levels are strongly predictive for atherosclerotic events/death, but Klotho levels were not (86). With regard to vascular calcification, in ESRD patients, Meuwese et al. did not find an association between soluble Klotho levels and CAC score (87), while Cai et al. find no association in women, a positive association in with AVC in men, and a borderline negative association with CAC in men (88). Finally, in a recent study, Berchtold et al. report that there is no association in kidney transplantation patients between Klotho level and a score for renovascular hyaline and fibrous intimal lesions (89).
All in all, many studies find negative associations between Klotho levels and parameters of atherosclerotic or calcific vascular disease, which seems promising, but our current inability to assess whether these measurements are specific for Klotho limits our capacity for interpretation. Perhaps these data will retrospectively be rendered valuable and the general use of the IBL assay makes results more easily comparable across studies. However with all commercially available assays, the problem remains that specifics about the antibodies and their epitopes are unknown and while the IBL assay does appear to detect Klotho, its specificity for Klotho is not unequivocally clear and the technical issues with regard to sample processing are a serious concern. Therefore, until the enduring need for a reliable ELISA is resolved and these studies can be replicated, caution is to be exercised in interpreting serum Klotho data.
425 Ta bl e 5 . Se ru m Kl oth o le vel s i n vasc ul ar d isea se. Re fer en ce N Po pu lat ion Pa th ol ogy ELISA Con ce ntra tio n Correl at ion s Con clusi on o n CVD Se m ba et a l., 2011 1023 Gen er al po pul at io n, pre do m in an tly el der ly CV D ( CA D, HF, stro ke, or PAD) IBL 2 nd te rti le 587 -7 69 pg/m L Ne gat iv e: ag e, al co ho l in ta ke, s m oki ng, MMSE , HDL ch ol es terol , tri gl yce rid es , CRP Lo w er pre val en ce o f CV D i n h igher Kl oth o le ve l t er til es Ki ta gawa et al. , 2013 114 CKD Arteri al s tif fn es s, en do th elia l dys func tio n, at he ro scl ero sis , vasc ul ar cal cif icati on IBL 616 [ 460 -756] p g/ m L Ne gat iv e: age, PT H, FE Pi Po siti ve : e GF R, 1,25D No d iffer en ce i n flo w -m edi ate d di la tio n (b ut signi fican t po siti ve corr el at ion w ith Kl ot ho le ve ls) . Lo w er Kl ot ho le ve ls in hi gher PWV. Lo w er Kl ot ho l ev el s i n pa tie nts w ith h igher IMT . Lo w er Kl oth o le ve ls p at ie nt s wi th a orti c cal cif icati on . Je on g et a l., 2013 120 (HIV) 20 (con tro l) HIV cART -rel at ed caro tid at he ro scl ero sis IBL 646 p g/m L (HIV, n o at he ro scl ero s is) 550 p g/m L (HIV, athe ro scl ero s is 708 p g/m L (c on tro l) Ne gat iv e: age, gl uco se , du ra tio n o f H IV in fe cti on Kl oth o i s an i nd ep en de nt ris k fact or f or cART -re lat ed ca ro tid a th ero sc le ro sis in H IV pa tie nt s/ Kai se r et a l., 2013 Shib at a et a l., 2013 100 Card iol ogy in pa tie nt s CV D, h ear t f ai lu re , hy pe rten sion IBL 434 [ 274 -595] p g/ m L - No ass oci ati on b etw ee n K lot ho le ve ls an d CVD Ka ra lli ed de e t al. , 2013 76 DM p at ie nts w ith H T an d alb um in ur ia Arteri al sti ffn es s IBL 431 p g/m L - No ass oci ati on b etwe en ch an ge in Kl oth o le ve ls a nd b ra ch ial p ul se p re ss ure o r ao rti c PWV
426 Se ile r et a l., 2014 444 CK D s ta ge 2 -4 Card iov as cu lar ev en ts /d ea th IBL 397 [ 326 -485] p g/ m L - Sol ub le Kl oth o l ev el s w er e n ot pre di cti ve fo r at he ro scl ero tic e ve nt s/d eat h. Su e t a l., 2014 134 Eld er ly hype rt en sive , non -hype rt en sive pat ie nts . N on -el de rly hype rt en sive s. Hy pe rten sion ? ? - Kl oth o l ev el s a re lo w er in h yp erten sion . Pa rk et a l., 2014 36 EH , RVH, con tro ls Hy pe rten sion IBL ~400 p g/ m L - Kl ot ho le vel s w er e l ower in h yp er te ns iv e pa tie nt s. Bu iten et a l., 2014 127 ESRD CAD IBL 460 [ 350 -620] p g/ m L Lo w er Kl oth o l ev el s w ere a ss oci ated w ith m ore C AD, al th ou gh n ot in de pe nd en tly M eu w es e et al. , 2015 98 ESRD p at ie nts Vas cu lar cal cif icati on IBL ~350 p g/ m L Po siti ve : f T3 Kl oth o l ev el s ar e n ot correl at ed w ith CA C score Cai e t a l. 2015 129 He m od ia ly sis pa tient s Vas cu lar cal cif icati on IBL 613 [ 379 -817] p g/ m L So lu ble K lo th o le ve ls w er e in ve rs ely correl ated w ith AAC s core Mo rita e t a l., 2015 157 CA D p at ie nts CA D a nd AVC IBL ~400 p g/ m L Kl oth o w as po siti ve ly ass oc iat ed w ith AVC in m en , w hile b ein g (b or de rlin e) n egat iv el y ass oci at ed w ith CA C in m en . No ass oci ati on s w ere fou nd in w om en . Di Lu llo et a l., 2015 100 Mi ld to m od era te C KD pa tie nt s AVC a nd M VC ? 895 (8 11 – 985) p g/ m L Kl oth o w as n egat iv el y a ss oci at ed w ith A VC . No re lat ion wi th M VC . Ke le s et a l., 2015 50 He al th y con tro ls Ath ero scl ero sis Cus ab io 459 p g/m L - Pa tient s w ith h igher Kl oth o l ev el s h ad hi gher FMD an d l ow er caro tid IMT . Ke le s et a l., 2015b 80 Ty pe 1 DM, he al thy con tro ls Ath ero scl ero sis Cus ab io 454 p g/m L - Pa tient s w ith h igher Kl oth o l ev el s h ad hi gher FMD an d l ow er car oti d IMT . Di ab eti cs h ad lo w er Kl oth o l ev el s. Ke le s et a l., 2016 34 He al th y con tro ls Coro nar y m icro vasc ul ar dys func tio n Cus ab io 482 p g/m L - Pa tient s w ith h igher Kl oth o le ve ls h ad a hi gher co ro nar y fl ow re se rv e.
427 Ab dal la h et al. , 2 016 88 ESRD p at ie nts Coro nar y an d caro tid at he ro scl ero sis IBL 474 p g/m L - Pa tient s w ith h igher Kl oth o le ve ls h ad a lowe r p re val enc e o f CA D an d h ad l owe r IMT v al ue s. Be rch to ld et al. , 2016 129 Ki dne y tra ns pl an ta tio n p ati ent s Vas cu lar l es ion s IBL 744 ± 246 pg /mL - No ass oci ati on b etw ee n K lot ho le ve ls an d tra ns pl an ta tio n-re lat ed v as cu lar les ion s. CV D, c ard iovascu lar di sea se; CAD , c or on ary ar ter y di sease ; HF, hea rt f ai lu re ; PA D, p er ip her al art er y di sea se; M M SE, mi ni -m en tal st at e exami nat ion ; H DL, h ig h-den sit y lip op rot ei n; C RP, C -re act iv e pro tei n; CKD , c hro ni c k id ney di sease ; PT H, p ar at hyr oi d hor m on e; FE Pi , f ract io na l exc re tion of p hosp ha te; eGFR, est imat ed gl om er ul ar fil tra tio n rat e; P WV, pu lse w ave vel oc ity; IMT , i nt ima -m ed ia t hi ck ness; HIV, hu man immu no de fic ien cy vi ru s; c ART, com bi ne d a nt i-re tro vi ral t he rap y; DM , d iab etes mel lit us; HT, h yp er te nsi on ; EH; esse nt ial h ype rt en sio n; RVH , re novascu lar hype rt en sion ; ESRD , en d-st ag e re nal d isease ; C AC, cor onar y art er y cal cif icat io n; AA C, ab dom in al ao rt ic cal cif icat io n; AV C, ao rt ic val ve cal cif icat io n; M VC , mi tral val ve cal cif icat ion ; FM D, fl ow -med ia te d d ilat ion .
428
Conclusion
The purpose of this review was to explore the potential of Klotho as a biomarker in vascular disease. We have assessed studies on Klotho gene variants, Klotho promoter methylation, Klotho mRNA expression, Klotho protein expression, and soluble Klotho levels, with regard to vascular diseases (mainly coronary artery disease and vascular calcification, as well as contributing risk factors, including arterial stiffness, hypertension, and dyslipidemia). We conclude that a number of gene variants, most notably KL-VS and G-395A, show promise as factors potentially associated with vascular diseases. Large replication studies in different populations and functional mutagenesis studies are to be performed to elucidate the effects of these gene variants. Approaches that employ the quantification of methylation, mRNA expression, or protein expression would be interesting to explore in the kidney or in PBMCs, as potential predictors for vascular disease. Finally, the most practical approach would certainly be to assess serum (or urine) levels of Klotho, for which a reliable ELISA is yet to be developed, the lack of which complicates the interpretation of current results. A role for Klotho as a biomarker in vascular disease can certainly be envisioned in the future as Klotho appears to be intricately linked to the pathogenesis of many vascular pathologies. Further defining the effects of gene variants and the reliable quantification of serum Klotho levels may be instrumental in this endeavour.
429
References
1. M. Kuro-o, Y. Matsumura, H. Aizawa, H. Kawaguchi, T. Suga, T. Utsugi, Y. Ohyama, M. Kurabayashi, T. Kaname, E. Kume, H. Iwasaki, A. Iida, T. Shiraki-Iida, S. Nishikawa, R. Nagai, Y. I. Nabeshima, Mutation of the mouse klotho gene leads to a syndrome resembling ageing. Nature. 390, 45-51 (1997).
2. H. Kurosu, M. Yamamoto, J. D. Clark, J. V. Pastor, A. Nandi, P. Gurnani, O. P. McGuinness, H. Chikuda, M. Yamaguchi, H. Kawaguchi, I. Shimomura, Y. Takayama, J. Herz, C. R. Kahn, K. P. Rosenblatt, M. Kuro-o, Suppression of aging in mice by the hormone Klotho. Science. 309, 1829-1833 (2005).
3. H. Kurosu, Y. Ogawa, M. Miyoshi, M. Yamamoto, A. Nandi, K. P. Rosenblatt, M. G. Baum, S. Schiavi, M. C. Hu, O. W. Moe, M. Kuro-o, Regulation of fibroblast growth factor-23 signaling by klotho. J. Biol. Chem. 281,
6120-6123 (2006).
4. A. Imura, A. Iwano, O. Tohyama, Y. Tsuji, K. Nozaki, N. Hashimoto, T. Fujimori, Y. Nabeshima, Secreted Klotho protein in sera and CSF: implication for post-translational cleavage in release of Klotho protein from cell membrane. FEBS Lett. 565, 143-147 (2004).
5. M. C. Hu, M. Shi, J. Zhang, T. Addo, H. J. Cho, S. L. Barker, P. Ravikumar, N. Gillings, A. Bian, S. S. Sidhu, M. Kuro-O, O. W. Moe, Renal Production, Uptake, and Handling of Circulating alphaKlotho. J. Am. Soc. Nephrol. 27, 79-90
(2016).
6. K. Lindberg, R. Amin, O. W. Moe, M. C. Hu, R. G. Erben, A. Ostman Wernerson, B. Lanske, H. Olauson, T. E. Larsson, The kidney is the principal organ mediating klotho effects. J. Am. Soc. Nephrol. 25, 2169-2175 (2014).
7. M. C. Hu, M. Shi, J. Zhang, H. Quinones, C. Griffith, M. Kuro-o, O. W. Moe, Klotho deficiency causes vascular calcification in chronic kidney disease. J. Am. Soc. Nephrol. 22, 124-136 (2011).
8. A. S. Go, G. M. Chertow, D. Fan, C. E. McCulloch, C. Y. Hsu, Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N. Engl. J. Med. 351, 1296-1305 (2004).
9. X. Zhou, K. Chen, H. Lei, Z. Sun, Klotho gene deficiency causes salt-sensitive hypertension via monocyte chemotactic protein-1/CC chemokine receptor 2-mediated inflammation. J. Am. Soc. Nephrol. 26, 121-132
(2015).
10. X. Zhou, K. Chen, Y. Wang, M. Schuman, H. Lei, Z. Sun, Antiaging Gene Klotho Regulates Adrenal CYP11B2 Expression and Aldosterone Synthesis. J. Am. Soc. Nephrol. 27, 1765-1776 (2016).
11. T. Shimada, Y. Takeshita, T. Murohara, K. Sasaki, K. Egami, S. Shintani, Y. Katsuda, H. Ikeda, Y. Nabeshima, T. Imaizumi, Angiogenesis and vasculogenesis are impaired in the precocious-aging klotho mouse. Circulation. 110,
1148-1155 (2004).
12. Y. Saito, T. Yamagishi, T. Nakamura, Y. Ohyama, H. Aizawa, T. Suga, Y. Matsumura, H. Masuda, M. Kurabayashi, M. Kuro-o, Y. Nabeshima, R. Nagai, Klotho protein protects against endothelial dysfunction. Biochem. Biophys.
430
13. T. Kusaba, M. Okigaki, A. Matui, M. Murakami, K. Ishikawa, T. Kimura, K. Sonomura, Y. Adachi, M. Shibuya, T. Shirayama, S. Tanda, T. Hatta, S. Sasaki, Y. Mori, H. Matsubara, Klotho is associated with VEGF receptor-2 and the transient receptor potential canonical-1 Ca2+ channel to maintain endothelial integrity. Proc. Natl. Acad. Sci.
U. S. A. 107, 19308-19313 (2010).
14. K. Chen, X. Zhou, Z. Sun, Haplodeficiency of Klotho Gene Causes Arterial Stiffening via Upregulation of Scleraxis Expression and Induction of Autophagy. Hypertension. 66, 1006-1013 (2015).
15. T. B. Tucker Zhou, G. D. King, C. Chen, C. R. Abraham, Biochemical and functional characterization of the klotho-VS polymorphism implicated in aging and disease risk. J. Biol. Chem. 288, 36302-36311 (2013).
16. D. E. Arking, A. Krebsova, M. S. Macek, M. Macek Jr, A. Arking, I. S. Mian, L. Fried, A. Hamosh, S. Dey, I. McIntosh, H. C. Dietz, Association of human aging with a functional variant of klotho. Proc. Natl. Acad. Sci. U. S.
A. 99, 856-861 (2002).
17. D. E. Arking, D. M. Becker, L. R. Yanek, D. Fallin, D. P. Judge, T. F. Moy, L. C. Becker, H. C. Dietz, KLOTHO allele status and the risk of early-onset occult coronary artery disease. Am. J. Hum. Genet. 72, 1154-1161 (2003).
18. A. F. Low, C. J. O'Donnell, S. Kathiresan, B. Everett, C. U. Chae, S. Y. Shaw, P. T. Ellinor, C. A. MacRae, Aging syndrome genes and premature coronary artery disease. BMC Med. Genet. 6, 38 (2005).
19. I. J. Deary, S. E. Harris, H. C. Fox, C. Hayward, A. F. Wright, J. M. Starr, L. J. Whalley, KLOTHO genotype and cognitive ability in childhood and old age in the same individuals. Neurosci. Lett. 378, 22-27 (2005).
20. H. K. Kim, B. H. Jeong, Lack of functional KL-VS polymorphism of the KLOTHO gene in the Korean population.
Genet. Mol. Biol. 39, 370-373 (2016).
21. J. Tavakkoly-Bazzaz, O. Tabatabaei-Malazy, M. Tajmir-Riahi, D. Javidi, M. Izadi, M. Shahrabi-Farahani, P. Amiri, M. M. Amoli, Absence of kl-vs variant of klotho gene in Iranian cardiac patients (comparison to the world populations). Dis. Markers. 31, 211-214 (2011).
22. E. J. Rhee, K. W. Oh, W. Y. Lee, S. Y. Kim, C. H. Jung, B. J. Kim, K. C. Sung, B. S. Kim, J. H. Kang, M. H. Lee, S. W. Kim, J. R. Park, The differential effects of age on the association of KLOTHO gene polymorphisms with coronary artery disease. Metabolism. 55, 1344-1351 (2006).
23. Y. Kim, J. H. Kim, Y. J. Nam, M. Kong, Y. J. Kim, K. H. Yu, B. C. Lee, C. Lee, Klotho is a genetic risk factor for ischemic stroke caused by cardioembolism in Korean females. Neurosci. Lett. 407, 189-194 (2006).
24. D. E. Arking, G. Atzmon, A. Arking, N. Barzilai, H. C. Dietz, Association between a functional variant of the KLOTHO gene and high-density lipoprotein cholesterol, blood pressure, stroke, and longevity. Circ. Res. 96,
412-418 (2005).
25. L. Invidia, S. Salvioli, S. Altilia, M. Pierini, M. P. Panourgia, D. Monti, F. De Rango, G. Passarino, C. Franceschi, The frequency of Klotho KL-VS polymorphism in a large Italian population, from young subjects to centenarians, suggests the presence of specific time windows for its effect. Biogerontology. 11, 67-73 (2010).
26. D. B. Dubal, J. S. Yokoyama, L. Zhu, L. Broestl, K. Worden, D. Wang, V. E. Sturm, D. Kim, E. Klein, G. Q. Yu, K. Ho, K. E. Eilertson, L. Yu, M. Kuro-o, P. L. De Jager, G. Coppola, G. W. Small, D. A. Bennett, J. H. Kramer, C. R. Abraham, B. L. Miller, L. Mucke, Life extension factor klotho enhances cognition. Cell. Rep. 7, 1065-1076 (2014).
